Secure mobile code computing in distributed remote environment

Mobile agent computing is an increasingly important paradigm which presents interesting new security challenges. Among these are the protection of host against malicious agents and protecting agents against hosts that execute agents. Especially of interest is the use of mobile agents in e-commerce settings, in which users release autonomous agents to roam around virtual stores and find desired products. In this research we investigated several schemes for dealing with assuring mapplets secure computation in a possible hostile environment without the need of special firewall or intrusion detection safe hardware. In this investigation we focus on realistic schemes which are feasible to implement. The presented mobile computing scheme do not require that the foreign host to be totally trustworthy that provides the same assurance as if the computation was carried using a traditional secure client/server model. The presented dynamical scheme is based on distribution of the algorithm on different hosts and making use of that the transactions has shared secrets by using the (k, n) secret-sharing threshold scheme. One of the positive outcomes of the current scheme is that it is distributed denial of service resistant system. The current proposed scheme presents a novel study of the tradeoff between the security and communication in remote computation. This formal study allows us to take the first step toward understanding the problem, indicating the significance of the various cryptographic assumptions, and as a result of the development of improved remote mobile computation schemes.